include tool calls

* Add type overloads for chat() method in _client.py

* Overloading

* Fix Overload Overlap

* Fix async chat

* Lint

* Reverse

---------

Co-authored-by: Simon Ottenhaus <simon.ottenhaus@kenbun.de>

README.md

@@ -2,24 +2,6 @@
 
 The Ollama Python library provides the easiest way to integrate Python 3.8+ projects with [Ollama](https://github.com/ollama/ollama).
 
-## Prerequisites
-
-You need to have a local ollama server running to be able to continue. To do this:
-
-- Download: https://ollama.com/
-- Run an LLM: https://ollama.com/library
-    - Example: `ollama run llama2`
-    - Example: `ollama run llama2:70b`
-
-Then:
-
-```sh
-curl https://ollama.ai/install.sh | sh
-ollama serve
-```
-
-Next you can go ahead with `ollama-python`.
-
 ## Install
 
 ```sh

examples/ps/main.py

@@ -0,0 +1,31 @@
+from ollama import ps, pull, chat
+
+response = pull('mistral', stream=True)
+progress_states = set()
+for progress in response:
+  if progress.get('status') in progress_states:
+    continue
+  progress_states.add(progress.get('status'))
+  print(progress.get('status'))
+
+print('\n')
+
+response = chat('mistral', messages=[{'role': 'user', 'content': 'Hello!'}])
+print(response['message']['content'])
+
+print('\n')
+
+response = ps()
+
+name = response['models'][0]['name']
+size = response['models'][0]['size']
+size_vram = response['models'][0]['size_vram']
+
+if size == size_vram:
+  print(f'{name}: 100% GPU')
+elif not size_vram:
+  print(f'{name}: 100% CPU')
+else:
+  size_cpu = size - size_vram
+  cpu_percent = round(size_cpu / size * 100)
+  print(f'{name}: {cpu_percent}% CPU/{100 - cpu_percent}% GPU')

ollama/_client.py

@@ -11,7 +11,7 @@ from copy import deepcopy
 from hashlib import sha256
 from base64 import b64encode, b64decode
 
-from typing import Any, AnyStr, Union, Optional, Sequence, Mapping, Literal
+from typing import Any, AnyStr, Union, Optional, Sequence, Mapping, Literal, overload
 
 import sys
 
@@ -97,6 +97,40 @@ class Client(BaseClient):
   ) -> Union[Mapping[str, Any], Iterator[Mapping[str, Any]]]:
     return self._stream(*args, **kwargs) if stream else self._request(*args, **kwargs).json()
 
+  @overload
+  def generate(
+    self,
+    model: str = '',
+    prompt: str = '',
+    system: str = '',
+    template: str = '',
+    context: Optional[Sequence[int]] = None,
+    stream: Literal[False] = False,
+    raw: bool = False,
+    format: Literal['', 'json'] = '',
+    images: Optional[Sequence[AnyStr]] = None,
+    tools: Optional[Sequence[Any]] = None,
+    options: Optional[Options] = None,
+    keep_alive: Optional[Union[float, str]] = None,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  def generate(
+    self,
+    model: str = '',
+    prompt: str = '',
+    system: str = '',
+    template: str = '',
+    context: Optional[Sequence[int]] = None,
+    stream: Literal[True] = True,
+    raw: bool = False,
+    format: Literal['', 'json'] = '',
+    images: Optional[Sequence[AnyStr]] = None,
+    tools: Optional[Sequence[Any]] = None,
+    options: Optional[Options] = None,
+    keep_alive: Optional[Union[float, str]] = None,
+  ) -> Iterator[Mapping[str, Any]]: ...
+
   def generate(
     self,
     model: str = '',
@@ -108,6 +142,7 @@ class Client(BaseClient):
     raw: bool = False,
     format: Literal['', 'json'] = '',
     images: Optional[Sequence[AnyStr]] = None,
+    tools: Optional[Sequence[Any]] = None,
     options: Optional[Options] = None,
     keep_alive: Optional[Union[float, str]] = None,
   ) -> Union[Mapping[str, Any], Iterator[Mapping[str, Any]]]:
@@ -136,6 +171,7 @@ class Client(BaseClient):
         'stream': stream,
         'raw': raw,
         'images': [_encode_image(image) for image in images or []],
+        'tools': tools or [],
         'format': format,
         'options': options or {},
         'keep_alive': keep_alive,
@@ -143,6 +179,28 @@ class Client(BaseClient):
       stream=stream,
     )
 
+  @overload
+  def chat(
+    self,
+    model: str = '',
+    messages: Optional[Sequence[Message]] = None,
+    stream: Literal[False] = False,
+    format: Literal['', 'json'] = '',
+    options: Optional[Options] = None,
+    keep_alive: Optional[Union[float, str]] = None,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  def chat(
+    self,
+    model: str = '',
+    messages: Optional[Sequence[Message]] = None,
+    stream: Literal[True] = True,
+    format: Literal['', 'json'] = '',
+    options: Optional[Options] = None,
+    keep_alive: Optional[Union[float, str]] = None,
+  ) -> Iterator[Mapping[str, Any]]: ...
+
   def chat(
     self,
     model: str = '',
@@ -209,6 +267,22 @@ class Client(BaseClient):
       },
     ).json()
 
+  @overload
+  def pull(
+    self,
+    model: str,
+    insecure: bool = False,
+    stream: Literal[False] = False,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  def pull(
+    self,
+    model: str,
+    insecure: bool = False,
+    stream: Literal[True] = True,
+  ) -> Iterator[Mapping[str, Any]]: ...
+
   def pull(
     self,
     model: str,
@@ -231,6 +305,22 @@ class Client(BaseClient):
       stream=stream,
     )
 
+  @overload
+  def push(
+    self,
+    model: str,
+    insecure: bool = False,
+    stream: Literal[False] = False,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  def push(
+    self,
+    model: str,
+    insecure: bool = False,
+    stream: Literal[True] = True,
+  ) -> Iterator[Mapping[str, Any]]: ...
+
   def push(
     self,
     model: str,
@@ -253,6 +343,26 @@ class Client(BaseClient):
       stream=stream,
     )
 
+  @overload
+  def create(
+    self,
+    model: str,
+    path: Optional[Union[str, PathLike]] = None,
+    modelfile: Optional[str] = None,
+    quantize: Optional[str] = None,
+    stream: Literal[False] = False,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  def create(
+    self,
+    model: str,
+    path: Optional[Union[str, PathLike]] = None,
+    modelfile: Optional[str] = None,
+    quantize: Optional[str] = None,
+    stream: Literal[True] = True,
+  ) -> Iterator[Mapping[str, Any]]: ...
+
   def create(
     self,
     model: str,
@@ -386,6 +496,38 @@ class AsyncClient(BaseClient):
     response = await self._request(*args, **kwargs)
     return response.json()
 
+  @overload
+  async def generate(
+    self,
+    model: str = '',
+    prompt: str = '',
+    system: str = '',
+    template: str = '',
+    context: Optional[Sequence[int]] = None,
+    stream: Literal[False] = False,
+    raw: bool = False,
+    format: Literal['', 'json'] = '',
+    images: Optional[Sequence[AnyStr]] = None,
+    options: Optional[Options] = None,
+    keep_alive: Optional[Union[float, str]] = None,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  async def generate(
+    self,
+    model: str = '',
+    prompt: str = '',
+    system: str = '',
+    template: str = '',
+    context: Optional[Sequence[int]] = None,
+    stream: Literal[True] = True,
+    raw: bool = False,
+    format: Literal['', 'json'] = '',
+    images: Optional[Sequence[AnyStr]] = None,
+    options: Optional[Options] = None,
+    keep_alive: Optional[Union[float, str]] = None,
+  ) -> AsyncIterator[Mapping[str, Any]]: ...
+
   async def generate(
     self,
     model: str = '',
@@ -431,6 +573,28 @@ class AsyncClient(BaseClient):
       stream=stream,
     )
 
+  @overload
+  async def chat(
+    self,
+    model: str = '',
+    messages: Optional[Sequence[Message]] = None,
+    stream: Literal[False] = False,
+    format: Literal['', 'json'] = '',
+    options: Optional[Options] = None,
+    keep_alive: Optional[Union[float, str]] = None,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  async def chat(
+    self,
+    model: str = '',
+    messages: Optional[Sequence[Message]] = None,
+    stream: Literal[True] = True,
+    format: Literal['', 'json'] = '',
+    options: Optional[Options] = None,
+    keep_alive: Optional[Union[float, str]] = None,
+  ) -> AsyncIterator[Mapping[str, Any]]: ...
+
   async def chat(
     self,
     model: str = '',
@@ -498,6 +662,22 @@ class AsyncClient(BaseClient):
 
     return response.json()
 
+  @overload
+  async def pull(
+    self,
+    model: str,
+    insecure: bool = False,
+    stream: Literal[False] = False,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  async def pull(
+    self,
+    model: str,
+    insecure: bool = False,
+    stream: Literal[True] = True,
+  ) -> AsyncIterator[Mapping[str, Any]]: ...
+
   async def pull(
     self,
     model: str,
@@ -520,6 +700,22 @@ class AsyncClient(BaseClient):
       stream=stream,
     )
 
+  @overload
+  async def push(
+    self,
+    model: str,
+    insecure: bool = False,
+    stream: Literal[False] = False,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  async def push(
+    self,
+    model: str,
+    insecure: bool = False,
+    stream: Literal[True] = True,
+  ) -> AsyncIterator[Mapping[str, Any]]: ...
+
   async def push(
     self,
     model: str,
@@ -542,6 +738,26 @@ class AsyncClient(BaseClient):
       stream=stream,
     )
 
+  @overload
+  async def create(
+    self,
+    model: str,
+    path: Optional[Union[str, PathLike]] = None,
+    modelfile: Optional[str] = None,
+    quantize: Optional[str] = None,
+    stream: Literal[False] = False,
+  ) -> Mapping[str, Any]: ...
+
+  @overload
+  async def create(
+    self,
+    model: str,
+    path: Optional[Union[str, PathLike]] = None,
+    modelfile: Optional[str] = None,
+    quantize: Optional[str] = None,
+    stream: Literal[True] = True,
+  ) -> AsyncIterator[Mapping[str, Any]]: ...
+
   async def create(
     self,
     model: str,

ollama/_types.py

@@ -52,6 +52,8 @@ class GenerateResponse(BaseGenerateResponse):
   context: Sequence[int]
   'Tokenized history up to the point of the response.'
 
+  tool_calls: Sequence[Any]
+  'List of tool calls made by the model.'
 
 class Message(TypedDict):
   """